http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
상관계수를 이용한 아다부스트 분류 기반의 도로에서 자전거 검출
이영학(Yeung-Hak Lee),고주영(Joo-Young Ko),윤상훈(Sang-Hun Yoon),노태문(Tae-Moon Roh),심재창(Jae-Chang Shim) 한국정보기술학회 2011 한국정보기술학회논문지 Vol.9 No.4
In this paper, we define to recognize the person riding a bike using correlation coefficient. This correlation coefficient between two variables, in which one is the person riding a bike and other is its background, can represent correlation relation. First, we extract edge vectors using Histogram of Oriented Gradients(HOG) which includes gradient information and differential magnitude as cell based. And then, the value, which is calculated by the correlation coefficient between the area of each cell and one of bike, can be used as the weighting factor in process for normalizing the HOG cell. This paper applied the Adaboost algorithm to make a strong classification from weak classification. In this experiment, we can get the result that the detection rate of the proposed method is higher than that of the traditional method.
셀 투영 벡터와 곡률의 연속법에 기반한 아다부스트 알고리즘을 이용한 보행자 인식
이영학(Yeung-Hak Lee),고주영(Jooyoung Ko),윤상훈(Sang Hun Yoon),노태문(Tae Moon Roh),심재창(Jaechang Shim) 한국정보기술학회 2011 한국정보기술학회논문지 Vol.9 No.1
This article presents advanced algorithm to recognize pedestrian and non-pedestrian on input images by using double-staged cascade method based on new feature vectors. Here we extracted two new feature vectors : cell-projection and curvature-HOG. The curvature is based on Histograms of Oriented Gradients (HOG) mechanism for point feature and the cell-projection is for well presented local area feature. By using AdaBoost algorithm we can recognize strong classification from weak classifications. For this we have two stages here: In the second stage, if it is not recognized pedestrian and non-pedestrian we can go for second stage. In the second stage, proposed system used another feature and strong classification non-recognized input image. For suggested algorithm, cascade method using the second stage AdaBoost algorithm, we obtain higher recognition rate than the other traditional methods for pedestrian and non-pedestrian.
지능형 휠체어 적용을 위한 기울기 히스토그램의 상관계수를 이용한 도로위의 이륜차 인식
김범국,박상희,이영학,이강화,Kim, Bum-Koog,Park, Sang-Hee,Lee, Yeung-Hak,Lee, Gang-Hwa 대한의용생체공학회 2011 의공학회지 Vol.32 No.4
This article describes a new recognition algorithm using correlation coefficient for intelligent wheelchair to avoid collision for elderly or disabled people. The correlation coefficient can be used to represent the relationship of two different areas. The algorithm has three steps: Firstly, we extract an edge vector using the Histogram of Oriented Gradients(HOG) which includes gradient information and unique magnitude for each cell. From this result, the correlation coefficients are calculated between one cell and others. Secondly, correlation coefficients are used as the weighting factors for normalizing the HOG cell. And finally, these features are used to classify or detect variable and complicated shapes of two wheelers using Adaboost algorithm. In this paper, we propose a new feature vectors which is calculated by weighted cell unit to classify with multiple view-based shapes: frontal, rear and side views($60^{\circ}$, $90^{\circ}$ and mixed angle). Our experimental results show that two wheeler detection system based on a proposed approach leads to a higher detection accuracy than the method using traditional features in a similar detection time.
지능형 휠체어 적용을 위해 Haar-like의 기울기 특징을 이용한 아다부스트 알고리즘 기반의 보행자 인식
이상훈,박상희,이영학,서희돈,Lee, Sang-Hun,Park, Sang-Hee,Lee, Yeung-Hak,Seo, Hee-Don 대한의용생체공학회 2010 의공학회지 Vol.31 No.6
In this paper, we suggest an advanced algorithm, to recognize pedestrian/non-pedestrian using differential haar-like feature, which applies Adaboost algorithm to make a strong classification from weak classifications. First, we extract two feature vectors: horizontal haar-like feature and vertical haar-like feature. For the next, we calculate the proposed feature vector using differential haar-like method. And then, a strong classification needs to be obtained from weak classifications for composite recognition method using the differential area of horizontal and vertical haar-like. In the proposed method, we use one feature vector and one strong classification for the first stage of recognition. Based on our experiment, the proposed algorithm shows higher recognition rate compared to the traditional method for the pedestrian and non-pedestrian.
등고선 영역에 대한 곡률 성분의 기울기 히스토그램을 이용한 3차원 얼굴 인식
김범국(Bum-Kook Kim),이영학(Yeung-Hak Lee),김태선(Tae-Sun Kim) 한국정보기술학회 2009 한국정보기술학회논문지 Vol.7 No.6
The curvatures of face and the face-shape using the depth information contain the most important personal facial information. In this paper, we develop a method for recognizing the range face images by combining the multiple face regions with face curvature HOG. For the proposed approach, the first step tries to find the nose tip from the extracted face area and has to take into consideration of the orientated frontal posture to normalize. And then, we calculate the curvature features: principal curvature, gaussian curvature, and mean curvature for each region. The second step of approach concerns the application of depth information and HOG. In the experimental results, using over the depth threshold value 45 (DT45) show the highest recognition rate among the regions, and the principal curvatures achieve 97.6% recognition rate, incase of curvature HOG.
독립된 특징의 캐스케이드 방법을 이용한 아다부스트 알고리즘 기반 보행자 인식
이상훈(Sang-Hun Lee),이영학(Yeung-Hak Lee),김태선(Tae-Sun Kim),서희돈(Hee-Don Seo) 한국정보기술학회 2010 한국정보기술학회논문지 Vol.8 No.8
We integrate the cascade method approach with independent feature to achieve accurate pedestrian/non-pedestrian recognition system, which applies Adaboost algorithm to make a strong classification from weak classifications. The features used in our system are Histogram of Oriented Gradients(HOG) which includes gradient information and basic Haar-like features which includes horizontal and vertical characteristic information. And then, a strong classification needs to be obtained from weak classifications for composite recognition method from each feature vectors. In the proposed method, we use one feature vector and one strong classification for the first stage of recognition. For the recognition-failed image, the other feature and strong classification will be used for the second stage of recognition. Experiment results on two common dataset and comparisons with traditional methods are given. Based on our experiment, the proposed algorithm shows higher recognition rate compared to the previous methods for the pedestrian and non-pedestrian.
정윤주(Yunju Jeong),이영학(Yeung-Hak Lee),이스라필 안사리(Israfil Ansari),이철희(Cheol-Hee Lee) 한국전기전자학회 2020 전기전자학회논문지 Vol.24 No.4
말벌 종은 모양이 매우 유사하기 때문에 비전문가가 분류하기 어렵고, 객체의 크기가 작고 빠르게 움직이기 때문에 실시간으로 탐지하여 종을 분류하는 것은 더욱 어렵다. 본 논문에서는 바운딩 박스를 이용한 딥러닝 알고리즘을 기반으로 말벌 종을 실시간으로 분류하는 시스템을 개발하였다. 훈련 영상의 레이블링 작업 시 바운딩 박스 안에 포함되는 배경 영역을 최소화하기 위하여 말벌의 머리와 몸통 부분만을 선택하는 방법을 제안한다. 또한 실시간으로 말벌을 탐지하고 그 종을 분류할 수 있는 최선의 알고리즘을 찾기 위하여 기존의 바운딩 박스 기반 객체 인식 알고리즘들을 실험을 통하여 비교한다. 실험 결과 컨볼루션 레이어의 활성함수로 mish 함수를 적용하고, 객체 검출 블록 전에 공간집중모듈(Spatial Attention Module, SAM)을 적용한 YOLOv4 모델을 사용하여 말벌 영상을 테스트한 경우 평균 97.89%의 정밀도(Precision)와 98.69%의 재현율(Recall)을 나타내었다. The hornet species are so similar in shape that they are difficult for non-experts to classify, and because the size of the objects is small and move fast, it is more difficult to detect and classify the species in real time. In this paper, we developed a system that classifies hornets species in real time based on a deep learning algorithm using a boundary box. In order to minimize the background area included in the bounding box when labeling the training image, we propose a method of selecting only the head and body of the hornet. It also experimentally compares existing boundary box-based object recognition algorithms to find the best algorithms that can detect wasps in real time and classify their species. As a result of the experiment, when the mish function was applied as the activation function of the convolution layer and the hornet images were tested using the YOLOv4 model with the Spatial Attention Module (SAM) applied before the object detection block, the average precision was 97.89% and the average recall was 98.69%.